#ISMEianWrites
Quantitative Data Analysis for Beginners in Business Research
Dr. SHAMPA NANDI,
Professor–Marketing,
ISME
“The essence of Mathematics is not to
make simple things complicated but to
make complicated things simple”—S.
Gudder
At the initial stage of research, researchers
in business or management area often face
a dilemma; which is first—theories/research
concepts, statistics or software like SPSS, Excel.
To be a systematic and good researcher, a person
should have basic research concepts, understand
statistical techniques and its assumptions and
finally make wise use of software. Most of the
time all these go hand in hand and error or
lack of understanding in one area often blocks
the progress in the other part. This article is
an attempt to discuss the basics of research
emphasising quantitative analysis part of any
business research. Topics like sample size,
sampling method, data collection procedure and
qualitative data analysis are not discussed here.
Research in Business or Management area deals
with the intangible latent concepts, attitudes
and behavioural aspect, therefore most of these
concepts are difficult to quantify. Suppose a re-
searcher wants to measure the Servqual dimen-
sions and its impact on customers (patients) sat-
isfaction and customer loyalty. Except “Tangible”
all the other dimensions of Servqual (Reliability,
responsiveness, assurance and empathy) are in-
tangible in nature. Therefore, a structured ques-
tionnaire has to be prepared by the researcher
to obtain the attitude of the customers on these
Servqual dimension in a quantifiable manner.
Question arises about the reliability and validi-
ty of the research instrument or questionnaire.
Reliability is nothing but the consistency of the
measurement instrument whereas validity meas-
ures whether the instrument is measuring what
the researcher intends to measure.
It is advised to do a pilot testing to check the
reliability of the questionnaire. Most commonly
“Cronbach Alpha” test is used to measure the re-
liability. If “Cronbach Alpha” value is more than
0.7 then it shows questionnaire is reliable. For va-
lidity of the research instrument, different types
of validity are measured. Mostly in business re-
search content validity, concurrent validity, pre-
dictive validity, convergent validity, nomological
validity, and discriminant validity are considered.
Unfortunately very common direct techniques are
not available to measure validity. Subject experts
can be consulted for questionnaire preparation
to enhance content validity, Cronbach alpha also
helps to measure the validity of the question-
naire to a certain extent. Structural equation
modelling technique is the most commonly used
for measuring the validity of any model. As a be-
ginner it is better to start with a proven set of
measurement scale used by an earlier research-
er in similar area (available in reputed journals)
and then modify according to a current context.
The next part is how to prepare any question in
different types of scales- nominal, ordinal, inter-
val or ratio scale. Most of the research method
books cover the definitions of all the above men-
tioned scale. The question often arises in the
mind of the researcher—what is the significance
of framing a question in each type of scale. Ques-
tion can be framed in any types of scale depend-
ing on the researcher’s objective as well as the
techniques going to be used to analyse the data.
Let’s take an example—Question about an em-